Not that much work, I got a victim machine that is just some scripts, qemu and VNC. It boils down to a keypress to get a new instance activated and when the VNC connection terminates the image is killed. I still place too much trust in the browser..
Not that much work, I got a victim machine that is just some scripts, qemu and VNC. It boils down to a keypress to get a new instance activated and when the VNC connection terminates the image is killed. I still place too much trust in the browser..
Also, wow, that's really neat, having everything Just Work(TM) like that. If you use this setup on a regular basis (and it sounds like you do), I can't help but suggest ditching QEMU's VNC server and using TigerVNC (from inside the guest), which automatically resizes the remote virtual display (assuming the guest VM is Linux). Other options include going full crazy and setting up Xpra, a rootless VNC alternative that's a bit heavyweight, or the potentially interesting idea of "DISPLAY=host:0 chrome &" inside the guest (of course with lots of intricate fussing with security).
I can understand overtrusting the browser; sadly thanks to the W3C (which is now basically run by a bunch of corporations) and their hundreds of thousand-page specifications, browsers are impossible to audit. Besides Dillo (LOL) and Cargo (bookmarks to come back to in 1 year), NetSurf seems very interesting, and builds pretty quickly (after you figure out the build process :D:D) too.
Quick glance at netsurf makes it seem that it would not be that many hours of work to have it use arcan as A/V/I backend, hmm..
After the full-on derp that is WebRTC and WebGL, WebUSB will be the next round of "fun".
And NetSurf is designed to be able to output to a framebuffer (with a rudimentary builtin toolkit), yep. (AFAIK you can get it to build the framebuffer mode on top of straight Xlib, for testing that mode on Linux.) NetSurf doesn't do audio yet (no <video> or <audio> tag support for starters) but yeah, an arcan/durdan-based UI would be pretty cool! :P
WebGL wasn't a complete failure, it gave us http://acko.net/. WebRTC is fun though - I remember someone mentioning (it might've been on here) how they asked a WebRTC implementer why they specified arbitrary TCP/IP communication channels, and the person responded that you might want to use it to talk to your toaster. That functionality lets us have https://webtorrent.io/ though, so...
The potential worst of WebUSB: cloud-based hardware capture devices, where all your oscilloscope's processing happens in the cloud, and you'd better pay your subscription on time.
The potential best of WebUSB: crowdfunded/distributed USB protocol debugging and live sharing - plug in your device (possibly inline with a software-controllable on/off switch, for cold resets), point a webcam at said device, maybe attach probes to some test points, walk away, and it's been reverse-engineered by tomorrow morning, possibly for a fee, possibly for free. This would cover at least 50% of reverse-engineering. I'm not sure how viable it would be through the browser, though...
I say let the slightly clinically insane specifications guys go on reinventing all the wheels. They leave a wake of things that open and shut in their path that are just capable enough to actually be useful. \o/